1. Field
Apparatuses and methods consistent with exemplary embodiments relate to a video telephony system, an image display apparatus, a driving method of an image display apparatus, a method for generating a realistic image, and a non-transitory computer readable recording medium, and more particularly, to a video telephony system, an image display apparatus, a driving method of an image display apparatus, a method for generating a realistic image, and a non-transitory computer readable recording medium, which can protect privacy when a video call is made using an image display apparatus located at a fixed location, and also can provide a realistic video call when a user's location is changed.
2. Description of the Related Art
Various methods have been used for protecting privacy during a video call. These methods are related to extracting a user's appearance, combining it with a virtual background, and showing the image to the other user. To extract a user image, the related art method may be a method which separates the user and the background based on color information of a single color background through a related art two-dimensional (2D) camera, or through a differential image and a previous frame by learning the background in advance.
In addition, as a three-dimensional (3D) depth camera has been introduced in recent years, the background can be easily separated by extracting depth information, and thus a user image separated from the background is combined with a virtual image and shown in virtual reality, games, and video call.
In the case of a video call, a user may be reluctant to expose the inside of a house or a current place where a call is made, there is an increasing demand for a method of substituting a background of a caller with a virtual image for the purpose of protecting privacy and security. In addition, as the size and resolution of displays increases, there is a demand for a method of increasing a sense of reality to make users feel as if they really talk with each other face to face.
The related art method of extracting a user image through a 2D camera needs to use color information using a single color background or require a step of learning the background without a user in advance to model the background. However, when the single color background is used, a color value may be changed according to an illumination condition, and in this case, there is a problem that a user's appearance including a background is extracted. In addition, when the background is learned in advance, the background which is changed according to a user's location needs to be learned every time the background is changed.
On the other hand, the method of using 3D camera depth information can separate a user and a background, but there is a problem that the contour of the user is not clearly separated.
In addition, when the separated image of the other user and the virtual image are combined and displayed, the background of the other user is fixed even if the user moves, and thus it is difficult for the users to feel as if they talk with each other face to face.